Identification of vehicle axle loads from bridge responses using updated static component technique

The identification of the axle loads of a moving vehicle from bridge response is studied. The objective is to develop the weigh-in-motion (WIM) technique to estimate the vehicle weight from the bridge responses without any disturbance to the vehicle’s traveling speed. The least-squares method using conventional regularization is considered, since it can provide not only the static gross weight but also the dynamic axle loads of the vehicle. However, the accuracy of the identified results from the method greatly depends on an assigned regularization parameter, which is very sensitive to the vehicle–bridge properties such as the speed of the vehicle or the roughness amplitude of the bridge surface. Therefore, it is difficult for it to be precisely assigned in practice. To overcome this problem, regularization using the updated static component (USC) technique is proposed. The technique extracts the static components of the identified axle loads and leaves only the associated dynamic components to be identified. Then the iteration of the identification is imposed to improve the accuracy of the identified results. The numerical examples of identification of the axle loads of a two-axle vehicle moving on a simply-supported bridge under various speeds of the vehicle and surface roughness amplitudes of the bridge are considered. It is found that, employing the proposed USC technique, the required parameter for the regularization becomes much less sensitive. In addition, the identified results obtained from the technique are substantially accurate compared to those from the conventional regularization.